On Mon, May 20, 2002 at 01:27:39PM -0700, Mike Mestnik wrote: > Now I'd like to tackel another problem, Quake3 is blocked durring this pause > (sometimes a long pause) This could result in poor sound and more inportantly > keybord performance. I don't know how to fix this withought using fork. Since > fork is realy heavy id like to avoid it. Could maby the xserver help offload > these sleeps, it should be sleeping any way right?
Nod, but it wants to be blocked, doesn't it? I don't know about long, unless there's a bug, the frame rates I get in q3 suggest that function is getting hit typically 2 or more times the frame rate (for swaps and clears) I see where you've said it's an infinite loop, but if that happens you're probably dead anyway as the card has stopped processing and that implies a (possibly difficult to track down) bug somewhere else. Carry on and you run out of buffers and die there. Can we repair the damage? That doesn't seem to have been that successful in the past in terms of recovery / carrying on running the 3d app - others have more knowledge here. If anything, cleanly exiting from 3d and, if recovery is possible or the 2d side / keyboard and mouse aren't all broken, the 2d side trying to recover might work better? Failing that, finding the bugs that cause those hangs is the only real fix. Does changing RADEON_MAX_OUTSTANDING to 1 help with the lag? (If anything you're more likely to see worse interactive performance by not waiting here, try increasing RADEON_MAX_OUTSTANDING and navigating the q3 menus to see that) I'd have said ideally it should wait (read sleep) until it's done, yes. The problem is still knowing when it's done without polling the register otherwise you just sleep instead of spinning, exactly the same worse case a sleep/spin for a complete cycle when you could have returned. Nicer for other things trying to run though. I guess you could make this function use an ioctl() into the drm and get it to put you on a wait queue woken after the driver updates the register it's polling? Maybe others will chip in with their thoughts. -- Michael. _______________________________________________________________ Don't miss the 2002 Sprint PCS Application Developer's Conference August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
